Oregon Health & Science University

Computational Biologist 1

Oregon Health & Science University  •  Portland, OR (Hybrid)  •  1 hour ago
Apply
AI can make mistakes so check important info. Chat history is never stored.

Job Description

Department Overview

This position is a member of the Computational Biology Program at OHSU and is based in the QBI Lab of Dr. Young Hwan Chang (Biomedical Engineering, Knight Cancer Institute).

The Computational Biologist 1 will support original cancer research in three focus areas of the lab: (1) agent-based modeling and virtual tissue development, i.e., computer simulations of how cells interact within tumor tissue, built from and tested against real patient data; (2) virtual staining, i.e., deep learning models that predict multiplex molecular stains from routine H&E-stained tissue images; and (3) agentic AI workflows, i.e., large language model (LLM)-based agents that plan and carry out multi-step coding, analysis, and modeling tasks.

The work involves high-dimensional imaging and omic data, statistical modeling, and team-oriented software development following best practices such as literature currency, tool benchmarking, reproducible code, and code sharing through GitHub. Duties include collaboration with investigators and cross-core partners, code and data management, development of tools and pipelines, and support of the lab's discoveries and scholarly work.

Function/Duties of Position

Data Management & Pipelines

  • Wrangle, clean, structure and store lab data, including paired H&E and multiplex tissue images, spatial omics data, sample metadata, and processing provenance.
  • Implement and document scalable, reproducible workflows on high-performance computing (HPC) and GPU resources so that lab members and collaborators can re-use them.

Agentic AI Workflows & Tools

  • Develop and maintain agentic AI workflows in which LLM-based agents use software tools to automate multi-step tasks such as data processing, code generation, model building and calibration, and literature review.
  • Build, test, and publish open-source software using structured software engineering best practices, with human review and validation of agent outputs.

Virtual Staining

  • Train, benchmark, and validate deep learning models that predict multiplex immunofluorescence or other molecular stains from H&E images.
  • Register paired slides, measure prediction accuracy at the pixel, cell, and tissue level, and assess whether predicted stains support the same biological conclusions as measured stains.

Agent-Based Modeling, Virtual Tissue & Statistical Analysis

  • Develop, calibrate, and validate agent-based and statistical models of the tumor microenvironment, including virtual tissue models initialized from multiplex imaging and spatial omics data.
  • Perform parameter estimation, sensitivity analysis, and statistical analysis of imaging and omic data to support biological conclusions, manuscripts, and grant applications.
  • Other duties as assigned.

Required Qualifications

  • Master's Degree in Computational Biology or related field; OR
  • Bachelor's Degree in Computational Biology or related field AND 3 years of relevant experience.

Knowledge, Skills, and Abilities

  • Demonstrated knowledge of statistical modeling applied to biomedical data, including regression and generalized linear models, hypothesis testing, and model evaluation (e.g., cross-validation, goodness-of-fit, uncertainty estimates).
  • Demonstrated knowledge of agent-based modeling or other mechanistic/simulation-based models of biological systems, as shown through coursework projects, a thesis, or research work.
  • Demonstrated knowledge of deep learning for image analysis, including image-to-image or generative models, implemented in PyTorch or TensorFlow.
  • Demonstrated knowledge of agentic AI workflows: applying LLM-based coding or analysis agents (e.g., Claude Code, GitHub Copilot agent mode, LangGraph) to automate multi-step computational tasks, including checking and validating agent outputs.
  • Applied expertise in Python and/or R for scientific computing in a Linux/Unix environment, including version control with Git/GitHub.
  • Experience independently performing analysis and management of high-dimensional biological data (e.g., multiplex tissue imaging, H&E whole-slide images, spatial omics, single-cell or bulk sequencing), from data cleaning through interpretation of results.
  • Proven track record of producing reproducible, well-documented code and analysis pipelines, as shown by shared code repositories, publications, a thesis, or project reports.
  • Proven track record of managing multiple concurrent projects and meeting deadlines in a research or technical setting.
  • Demonstrated written and verbal communication of technical results to multidisciplinary audiences (e.g., presentations, manuscripts, reports).

Preferred Qualifications

  • Master's degree or PhD in Computational Biology, Biomedical Engineering, Biostatistics, Applied Mathematics, Physics, Computer Science, or a related quantitative field, with a thesis or research project focused on agent-based or mechanistic modeling, statistical modeling, deep learning for tissue image analysis (e.g., virtual staining), or LLM-based agentic AI systems.
  • Applied expertise in developing, calibrating, and validating agent-based models of multicellular or tissue-level systems (e.g., PhysiCell, CompuCell3D, Chaste, Mesa, or custom frameworks).
  • Demonstrated knowledge of virtual tissue development: building spatially resolved in silico tissue models that are initialized from, or compared against, experimental data (e.g., multiplex imaging, spatial transcriptomics, H&E).
  • Applied expertise in virtual staining: training and evaluating deep learning models that predict multiplex immunofluorescence, IHC, or other molecular stains from H&E or label-free images, including registration of paired slides and quantitative validation of predicted stains.
  • Proven track record of building agentic AI systems, such as LLM agents with tool use, multi-agent orchestration, or Model Context Protocol (MCP) servers, applied to scientific analysis or modeling (e.g., LLM-assisted construction or calibration of agent-based models).
  • Experience independently performing parameter estimation, sensitivity analysis, and uncertainty quantification for simulation models, including simulation-based inference or Bayesian calibration (e.g., approximate Bayesian computation).
  • Applied expertise in advanced statistical modeling, such as hierarchical/mixed-effects models, Bayesian inference, spatial statistics, or survival analysis.
  • Proven track record of combining machine learning or deep learning with mechanistic models (e.g., surrogate models/emulators, hybrid models).
  • Experience independently running large-scale model training, simulations, or parameter sweeps on HPC clusters, GPUs, or cloud platforms (e.g., SLURM, containers, Nextflow/Snakemake).
  • Demonstrated knowledge of cancer biology, the tumor microenvironment, and computational pathology.
  • Proven track record of peer-reviewed publications, preprints, or conference presentations in computational biology, computational pathology, or biological modeling.

Additional Details

Benefits

  • Healthcare for full-time employees covered 100% and 88% for dependents.
  • $50K of term life insurance provided at no cost to the employee.
  • Two separate above market pension plans to choose from.
  • Paid time off - 208 hours per year, prorated for part-time.
  • Extended illness bank - 64 hours per year, prorated for part-time.
  • 9 paid holidays per year.
  • Substantial Tri-Met and C-Tran discounts.
  • Employee Assistance Program.
  • Childcare service discounts.
  • Tuition reimbursement.
  • Employee discounts to local and major businesses.

Why apply to OHSU?

We are Oregon's only public academic health center. In addition to caring for patients, we lead groundbreaking research. We also train the next generation of health care professionals. As Portland's largest employer, we give you opportunities to learn and advance in a system of hospitals and clinics across Oregon and Southwest Washington. All are welcome. OHSU welcomes people of all ages, ethnicities, genders, national origins, religions and sexual orientations. We are striving to build an anti-racist, multicultural institution and encourage people with diverse backgrounds to apply. To request reasonable accommodation, contact askhr@ohsu.edu

Oregon Health & Science University

About Oregon Health & Science University

Industry
Unknown
Company Size
Unknown
Headquarters
Unknown
Year Founded
Unknown
Website
ohsu.edu
Social Media